back
user profile

raphlinus

13,685karma·1,846submissions·March 7, 2014
about
I do research on fundamental UI technology and 2D graphics, with a focus on Rust and fonts. @raph@mastodon.online
recent activity (1,846 total)
comment
So this is anecdote and not data, but I'm interacting a lot more with Quakers on Mastodon than I did on Twitter. Honestly, I don't think services need to be dumbed down so much, people can f…
3y ago·view thread
comment
TIL :)
3y ago·view thread
comment
I've also moved from Twitter to Mastodon, for the same reasons, and have a similar experience - the raw follower count is still a fraction as before, but a lot of interesting people I want to con…
3y ago·view thread
comment
I got "Install the NuGet package manager on a Mac" which I'm still not sure is ridiculously infeasible or the kind of hack somebody might actually manage to pull off. Definitely HN-wort…
3y ago·view thread
comment
Or, if you prefer Mastodon: https://mastodon.online/@kenshirriff@oldbytes.space/10941232... …
3y ago·view thread
comment
Yeah, the way I sometimes describe this is that Rust social media can be pretty toxic, but the actual Rust community is generally very friendly and supportive. The problem is that a lot of people conf…
3y ago·view thread
comment
The `fast_decode` as written is undefined behavior for some inputs (INT_MAX for example) because of the integer overflow in `2*x`. This can be fixed by casting to an unsigned int first. Also a warning…
3y ago·view thread
comment
What I think has happened is that the culture wars are taking over more and more topics. Not only the obvious stuff like race, LGBTQ+ rights, crime, and labor, but increasingly things that don't …
3y ago·view thread
comment
I think it's potentially in scope, but optional - there are other problems more urgent to solve, and then with more powerful compute and font technology, it's possible to go past TeX into mi…
3y ago·view thread
comment
How does it work? How does it compare to other published work, including piet-gpu? I'm seeing about 5ms for tiger on a Pixel 4, which is probably comparable as the iPad no doubt has a much strong…
3y ago·view thread
comment
I've been thinking of writing more. Doing real text layout is one of the tricky bits, and one that terminals sidestep because layout calculations are trivial as long as you narrow scope down to…
3y ago·view thread
comment
To some extent, the architecture of xi-editor was inspired by similar thoughts - not escape codes, but updates to the UI specified as JSON-RPC, and with one process handling the nitty-gritty of UI, th…
3y ago·view thread
comment
I was quite impressed with the FreieFarbe[1] presentation at Libre Graphics Meetup 2019. It is an open source service that actually does attempt to provide similar value as Pantone, not just the namin…
3y ago·view thread
comment
I love these kinds of things and use them in GPU programming, among other things. Things have changed in a variety of ways: population count and count-trailing-zeros are generally available as fast in…
3y ago·view thread
comment
> GPUs don't yet have a well-defined memory model Vulkan has a perfectly fine memory model, and it's non-optional in Vulkan 1.3. The rest of GPU land of course has some catching up to do.…
3y ago·view thread
comment
Also see the twofloat crate in Rust, which uses a pair of f64's to give double the number of significant digits as a standard f64. The linked docs point to a number of academic papers on the subj…
3y ago·view thread
comment
Some more wild shady spy activity against whistleblowers documented here: https://twitter.com/HindenburgRes/status/1575557272532811778 …
3y ago·view thread
comment
Perhaps the rule in the standard is simple - the compiler can arbitrarily round to finer precision than IEEE, but in practice it's complicated as the same code can behave quite differently depend…
3y ago·view thread
comment
The actual rules are very complicated. C allows greater precision for intermediate results but compilers are sometimes careful to stick to IEEE rounding. [1] contains a good general overview, and [2] …
3y ago·view thread
comment
In fact I do, Bellingcat has published extensive analysis[1] of this attack. [1]: https://www.bellingcat.com/news/2022/04/14/russias-kramators... …
3y ago·view thread
comment
Markdown parsing can be quite fast. I just benchmarked pulldown-cmark parsing all the Markdown in my blog (10k lines, 1MB) and writing HTML, and it does that in 14ms. Admittedly that's on an M1 M…
3y ago·view thread
comment
If by "controversy" you mean disinformation techniques designed to cast doubt on credible reports of war crimes, you are correct. See [1] for an analysis of this claim, including tracing it …
3y ago·view thread
comment
For now. Unfortunately, even if you have very fast drawing, there are plenty of other ways to make software slow. But high performance GUI software is the main thing I'm working on at the momen…
3y ago·view thread
comment
The main distinction I'd draw is whether you'd doing formal or informal reasoning. In the formal reasoning world, undefined behavior is mostly a good thing. On one side of the contract, it&#…
3y ago·view thread
comment
I vouched this. I have no idea whether it's true, but I think it's worth discussing and relevant to the topic. My first real introduction to fixed points was in Dijkstra's weakest preco…
3y ago·view thread
comment
A ton of stuff, arguably too many things. Right at the moment, my main focus is redoing fundamental 2D vector graphics geometry operations (stroking, path intersection) to make them more robust and hi…
3y ago·view thread
comment
Can you send me some screenshots? Very possibly you have some ClearType parameters set wrong on your system. We absolutely do test on 1080p (though I admit my work machines are all retina/4k at t…
3y ago·view thread
comment
Behdad had this result earlier: https://twitter.com/behdadesfahbod/status/958065026262183936 …
3y ago·view thread
comment
I find Neil Mitchell's categories of small[1], medium[2], and huge[3] build systems useful. Blaze is absolutely fantastic as a huge build system, as it can correctly specify the exact semantics o…
3y ago·view thread
comment
That's a reasonable choice and I do agree it has nice properties, for example (a << b) << c is equal to a << (b + c) (unless that latter addition overflows), but it also does pu…
3y ago·view thread